Confusion simplex coloured by Matthews Correlation Coefficient
- Mouse over the tetrahedron, then click and drag to change its orientation.
- Click on the text
Pos==20 to toggle that slice of the confusion matrix.
Confusion simplex coloured by Accuracy
- Mouse over the tetrahedron, then click and drag to change its orientation.
- Click on the text
Pos==20 to toggle that slice of the confusion matrix.
About
These interactive plotly visualisations show 3D projections of binary confusion matrices of size \(N=100\).
- Each plot depicts the value of a different confusion matrix performance metric.
- Each point corresponds to a unique confusion matrix and is coloured by the value of that metric.
- For reference, we label the four extreme points corresponding to all True Positives, (\(\mathrm{TP}=100\)), all False Negatives (\(\mathrm{FN}=100\)), etc., and connect those vertices to give an impression of the regular tetrahedral lattice (i.e., the 3-simplex) of the projected points.
- In total, there are \(\binom{100+4-1}{4-1}=176\,851\) different binary confusion matrices with of size 100.
- Rather than show all these, we have taken three slices through the lattice: from back to front, the rectangular lattices of points correspond to confusion matrices where \(\mathrm{Pos} = 20, 50, 90\), respectively.
- If you rotate the plots so that \(\mathrm{TP}\) is at the top, \(\mathrm{FN}\) is at the back left and \(\mathrm{FP}\) at the back right, the axes of these slices will be oriented like those of conventional ROC curves:
- \((\mathrm{TN},\mathrm{TP})\) are maximum at the top-left and minimum at the bottom-right;
- the diagonals from bottom-left to top-right correspond to confusion matrices from random classification.